Using a focused diversity measurement system can make a critical difference in the performance effectiveness of your diversity ROI metrics. In Part 1, I gave an overview of a five-step process to consider and highlighted some key background issues. In this segment (Part 2), I will detail each step of the five-step process including possible diversity metrics, formulas and suggestions. Although measuring diversity Return on Investment (DROI) impact is not an exact science, there are a number of valid techniques, tools and reliable methods for translating business performance gains into tangible financial results that C-suite executives will support and embrace. Let’s take a look at the specific steps I am recommending in detail.
Key steps to building an effective measurement system
Creating an effective diversity measurement system involves at least five critical steps:
1. Review the Strategic Business Plan
If diversity is to address the strategic business challenges of the organization, then it is imperative that the needs of the business are understood. The best place to start is the strategic business plan. The strategic business plan (or its equivalent) can offer a wealth of information regarding areas where diversity can make a difference or contribute. It is critical to examine the mission of the organization, key goals and objectives, core business strategies and tactics, marketing and sales targets, production or operational issues, recruitment and retention issues, succession plan challenges, productivity issues, legal compliance concerns, customer service challenges, globalization plans and the like. Each of these areas should be examined to determine where diversity could have a direct or indirect impact on the business. The results of this activity should form the basis for creating your diversity initiatives.
2. Formulate Research Questions
Know what you want to know. This involves formulating research questions that help give you answers for establishing baseline diversity measures or calculating change and impact. Useful approaches for formulating research questions include starting broad then narrow the focus. Using strategic business plan information as your guide, focus on helping the organization solve and/or prevent business problems as well as improving business operations. An example might include questions such as “What is the distribution and retention impact of females and minorities in management positions above the first level?”, “What is the diversity makeup of our customer base by product line and market share?”, “What factors make the difference for high productivity among diverse work teams and their impact on customer service ratings?”, or “What percentage of mothers return to work after maternity leave by diverse groupings compared to industry and majority group norms?”
These research questions should help align and link the diversity initiative with the strategic plan. If they are systemic, diversity initiatives will also proceed along several lines at the same time. Activities such as child-care services, flexi-time, flexi-place strategies, parental leave options, mentoring and others could be in action simultaneously. Let’s suppose that at the end of a two-year period, management notes a 45 percent decrease in turnover among female managers. The impact of this decrease is an identifiable value to the organization in lower recruiting and training costs. However, which of the above programs caused the improvement during that period?
Most of the time, diversity can take only partial credit for improvement. To help isolate the effects of an initiative from other factors that could have affected the result, you will need to go beyond standard control group analysis to use one or more techniques for isolating extraneous factors. Three techniques, in addition to control group methods, include:
• Trend-line analysis — Using a graph of mothers returning to work after maternity leave, for example, a line is drawn from current percentages to future percentages, assuming that the current trend would continue even without diversity initiatives. After the diversity intervention is implemented, post statistics are compared to the trend predicted on the trend line. It is reasonable to attribute any improvement over the trend-line prediction to the diversity intervention. It is not an exact process, but it does provide a reasonable estimate of this diversity intervention’s effectiveness.
• Forecasting — This approach is more analytical and mathematical than the trend-line. Instead of drawing a straight line, a linear equation is used to calculate a value of the anticipated performance improvement or impact. A linear model (such as y = ax+b) is appropriate when only one variable influences the result. When several variables intervene, it’s necessary to use sophisticated statistical models. Without them, forecasting is difficult to implement. Still, it can be an accurate predictor of performance variables without the diversity intervention, if the appropriate data and models are available.
• Employee and Supervisor Estimations — This approach involves asking employees and supervisors to determine how much performance improvement or environmental change is due to specific diversity interventions. Their actions have produced the improvements, so they should have some idea of how much improvement is due to their participation and use of the diversity initiative. Upper management will tend to find these reports credible because employees and supervisors are at the center of these improvements or changes. Employees’ and supervisors’ input can be obtained by asking questions such as “What percentage improvement can be attributed to the implementation of the ‘XYZ’ diversity initiative?” or “What is the basis of your estimation?”or “What degree of confidence do you have in your estimation?” or “What other individuals or groups could make an estimate?” or “What other factors do you think contributed to the improvement?” To be conservative, it is recommended to factor in a confidence level. For example, if an employee estimates that 50 percent of an improvement seen in turnover statistics among Hispanic employees is due to the minority career development network initiative but is only 70 percent confident about that estimate, multiply the confidence percentage by the improvement percentage and divide by 100, for a confidence level of 35 percent. Then multiply that figure by the amount of the improvement in order to isolate the portion attributed to the diversity intervention. To calculate return on investment (ROI), convert that portion to a monetary value.
3. Design the Study Methodology
This step in the process involves creating a formalized plan of action, which spells out how you will address each research question and the potential tools used. A complete plan should address questions such as “What existing measures can be adopted to measure these areas?” or “What benchmark or compatible measures are available?” or “What is the priority or order of measures by criticality level?”
4. Collect and Analyze Data
The data collection and analysis process requires the use of specific formulas and techniques designed to assess the research questions. These formulas and techniques might include calculations such as Cost per diversity hire, percent change in local and global customer diversity demographics, the Family of Measures Index, diversity hit rate, diversity training evaluation using the Kirkpatrick Model-Levels 1-5 surveys along with ROI calculations, etc. For example, to measure the distribution of performance and retention impact of females and minorities in management positions above the first level, you might use the following calculations:
• Percentage of Minorities and Women in Key Leadership Positions — Formula: PMW = (PF/LP) * 100 where: PMW = number of minorities and women in key leadership positions, PF = number of leadership positions filled with minorities and/or women (e.g., 19), LP = number of leadership positions (e.g., 65). Therefore, the percentage would calculate as PMW = (19/65)*100 or .2923*100 = 29.2 percent.
Changes in these numbers might signal forward or reverse trends in diverse workforce promotional changes.
• Diversity Hire Performance Impact — Formula: DHPI = DHPR+DHP+DHS/N where: DHPI = Diversity Hire Performance Impact, DHPR= average job performance rating of new diversity hires (e.g., 4 points out of a 5 point performance rating scale = 80 percent), DHP= Percentage of new diversity hires promoted within one year (e.g., 45 percent), DHS= Percentage of new diversity hires retained after one year (e.g., 90 percent), N= number of indicators used. Therefore, the percentage would calculate as DHPI = (80+45+90)/3 or (215)/3 = 71.7 percent. The resulting percentage is a relative value. You must determine whether this value represents high, medium or low performance impact for this hire group. This comparison can be based on historical comparisons, preset performance standards or objectives, or management mandates.
In order to have maximum impact in demonstrating diversity’s value to the bottom-line, it is vital that percentage calculations are converted to financial terms (dollars and cents) whenever possible. These percentages might convert to a preventative cost savings such as a turnover dollar saving, lower recruiting costs or a benefit such as increased customer satisfaction ratings, added customer purchase volume, improved employee morale and productivity (measured in increased units per hour, faster problem resolution using diverse teams which results in reduced cycle time, speed to market), etc.
For the most part, any object, issue, process or activity that can be described by observable variables is subject to measurement. Phenomena can be evaluated in five dimensions of measurement: cost, time, quantity, quality or frequency of occurrence.
The central issue in applying measurement to the diversity culture change process is to decide what is worth measuring and agree on the measure as a fair representation of progress and accountability. Which measurement tools you use will be determined by what you want measured and what results you want from the measurements. Direct measures are always identifiable by the fact that they measure some kind of cost. Indirect measures do not cover costs, but do describe some measure of time, quantity or quality.
Report card measures of diversity are like snapshots of the past. They can provide a historical reference to accomplishment or serve as milestones along the path to producing outcomes. Examples might be the number of diversity training sessions completed, cost per diversity trainee hour, turnover by performance level by gender by length of service.
5. Communicate and Monitor the Results
Reporting your diversity results is almost as important as producing the results. Regardless of the message, at least three general rules are important to remember:
• The communication must be targeted to specific audiences — The communication will be more efficient when it is designed for a specific group. The message can be tailored to the interest, needs and expectations of the intended audience. The length, content, details and slant will vary with the audience.
• The communication should be unbiased and always modest — Facts must be separated from fiction, and accurate statements must replace opinion. Some target audiences may view communication from the Diversity Department with skepticism and may look for biased information and opinions. Boastful statements will sometimes turn individuals off, and most of the content of the communication will be lost. Observable, believable facts carry more weight than extreme, sensational claims, although the claims may be needed to get initial attention.
• The communication must be consistent — The timing and content of the communication should be consistent with past practices. A special communication at an unusual time may create suspicion. When a particular group, such as top management, regularly receives communication, the information should continue even if the results are not good. If selected results are omitted, it might leave the impression that only good results are reported. Find out how you can become a regular participant and presenter in key departmental staff meetings. Encourage others to join your staff meetings to create strategic partnerships. Finally, be sure to develop good measurement monitoring practices. This will ensure a continuous feedback loop to meet the organization’s informational and development needs in a responsive manner.
I hope that this article will help you begin the journey of measuring your diversity results. You can use measurements to identify winning approaches and processes that work to transform the organization’s culture to a more inclusive environment. They can help you increase the value of your workforce diversity efforts by providing focus and demonstrating organizational impact. If full use of the diverse workforce is to be a reality in our lifetime, we must use every tool or resource available to fully monitor and communicate the effectiveness of this effort.Each step in the process logically builds on the previous step, which generates an evidenced-based framework that creates a “Best Practice” method for proving Diversity’s link to performance.
If you put these steps in operation, we want to hear about your experience and share it with others. Let us know about your success with the diversity ROI metrics process to help elevate the field.
